摘要
分析了典型的聚类算法及其适用范围,针对其处理Web点击流数据的不足,提出了一种用于Web点击流的增量挖掘的聚类算法WCSCluster,给出了相关定义及存储结构,并用实例说明了算法的运行过程.最后对比同类算法给出实验结果.实验结果表明该算法具有良好的性能,能够发现更优的簇.
After analyzing the classic clustering algorithms and its confine,an incremental mining clustering algorithm which is used in Web clickstreams field called WCSCluster is proposed.Relevant definitions and storage structure are given;running procedure is shown by a specific example.Experimental results are given comparing with congener algorithms.Experimental results demonstrate that the algorithm has good performance which can find superior clusters.
关键词
WEB使用挖掘
点击流
增量挖掘
聚类
Web usage mining
clickstreams
incremental mining
clustering
